SY,raintalk's post came at the right time.However one of the tests I've done several times was to try for audio anything that came my way for decades now.I will never forget how much I liked two single strands of 0.75mm which I removed from a speaker cable run and which I tried without any insulation.When I decided to dress them,only thing I could use werw shrinkable sleeves.So I did,while trying to retain the configuration the same,so not to change the LCR values at all by changing distance between the two strands.I did as close as possible as i could.I really felt nice untill it went back between the cd player and amp I was using at the time.I know,I know what you'll say but spend some time to try something similar.Mine went to the dustbin after two hours I let it abuse my music.Some time later a friend ordered some teflon sleeves and we did the same thing.I dont know for him but on my system it served nicely for 18months until I replaced them.
Panicos K said:
Resolution, depth, especially transient response indeed change a lot by complex loads since any amp has non-linear output. Amplifiers with deep feedback and without feedback are affected much more, probably that's why audiofiles so avare of cables that "sound", when in reality they affect the way output stages distort. Since audiofiles are not equal: some of them prefer less distortions, some of them prefer certain kind of distortions, no single receipt what cables are better exist. But it does not mean that cables do something magical. Or yes, magical, if to call magical everything what people don't understand. In this particular case they don't understand why some cables that capacitive load amplifier may "sound better". But if to remember that criteria is subjective, it is simple: capacitive load changing the way the amp distorts ad some "signature" that some people love.
wavebourn,resolution,depth as well as many other things that make music more believable,are contained in many good recordings and they change convinsingly as theese recordings become better.Still,if what you say,that depth and resolution are the result of some cables to effect the way amplifier output stages distort then it's fine with me.But you just said that speaker cables sound different.Planet 10?
Panicos K said:
Let me try... I'll defenitely wind them around a magnet to listen how do they sound... Later...
Now I am busy with my power cords blind-testing their sonic signatures.
Also, I am developing a revolutinary technology of fast-NOSing of modern tubes. As vintage bottles are flat on one side, the same way NOS tubes are magnetized by a magnetic field of the Earth... It is one of the deepest secrets of real AudioFeels. 😎
"Dream Loudspeaker"
The name says it all.
BTW, fat chance keeping Helium in any flexible tube. It seeps right thru - same pressure or not.
The name says it all.
BTW, fat chance keeping Helium in any flexible tube. It seeps right thru - same pressure or not.
panomaniac said:
I see a fortune to be made refilling speaker cables. For an extra fifty I'll top up the electrons. Maybe toss a couple protons your way if I'm in a good mood.
No, just trying to speak the same language...Panicos K said:
I NEVER said that cables SOUND. I said that well known physical parameters affect amplifiers changing character of their distortions. No miracles in loading amplifiers on complex loads. Miracles are result of different tastes: some people prefer less distortions, some people prefer distortions of a certain type, some people just have no own preferences, but believe Gurus. It is my explantion that corellates well with what I observe around me. All of them are called audiofiles.
However, you may have your own point of view that you like more than mine.
With the help of a friend, we did some tests on 3 interconnects because I did not have three commercial speaker cables. We did two sessions. The first one came out 3 correct identification out of 10 (additional 3 had two cables consistently reversed in identification), the second session was 5 correct identification out of 10.
The setup consisted of Hafler XL280 amp. Aboss HD-9000 multi format dvd player, peaker cables were Alpha Core Serpents, speakers were Augene Lullaby's. Since we forgot to bring the remote control to the test site, I had to listen at full volume which was much more than the speaker was designed for.
During the first session, I listened to each interconnect using various selected material just to get a feeling as to how they sounded in the test environment. After that we went directly into testing. I would request the source CD to be played, and after I made a decision, I would write it down and aske for cable change. My friend would choose the cable without me knowing what it was, it was not necessary to change, but the action of removing and connecting was required. After 10 identifications, we would compare notes.
During the second session, I again listend to all of the cables using only a few music sections until I was able to identify more specific characteristic that would help me identify the right interconnect more accurately. After this was established, we went right into testing the same way as in the first session. This time I shortened the listening time before each idenfication. As a result, the first 4 and the 6th identification were correct.
Based on this test we had concluded that:
1. The longer time you listen to each material, the more easily you get mixed up.
2. The further into the test, the more likely error will occur although mathematically they should be the same.
3. I felt my self getting more mixed up the further I get into the test.
4. With 3 cables the statistical success rate should be 1/3. But in reality, if the exact reverse identification counts in some way, the success rate was at least 50% for both sessions.
I think if we were in a more acoustically dead room, and the equipment were even better, the success rate would be even higher.
The setup consisted of Hafler XL280 amp. Aboss HD-9000 multi format dvd player, peaker cables were Alpha Core Serpents, speakers were Augene Lullaby's. Since we forgot to bring the remote control to the test site, I had to listen at full volume which was much more than the speaker was designed for.
During the first session, I listened to each interconnect using various selected material just to get a feeling as to how they sounded in the test environment. After that we went directly into testing. I would request the source CD to be played, and after I made a decision, I would write it down and aske for cable change. My friend would choose the cable without me knowing what it was, it was not necessary to change, but the action of removing and connecting was required. After 10 identifications, we would compare notes.
During the second session, I again listend to all of the cables using only a few music sections until I was able to identify more specific characteristic that would help me identify the right interconnect more accurately. After this was established, we went right into testing the same way as in the first session. This time I shortened the listening time before each idenfication. As a result, the first 4 and the 6th identification were correct.
Based on this test we had concluded that:
1. The longer time you listen to each material, the more easily you get mixed up.
2. The further into the test, the more likely error will occur although mathematically they should be the same.
3. I felt my self getting more mixed up the further I get into the test.
4. With 3 cables the statistical success rate should be 1/3. But in reality, if the exact reverse identification counts in some way, the success rate was at least 50% for both sessions.
I think if we were in a more acoustically dead room, and the equipment were even better, the success rate would be even higher.
OK, so you got nothing significantly different than random guessing. Chances are that you've got decent enough electronics to avoid having unwanted interactions with the load and indeed there's no difference. But we shall keep an open mind and do a check.
Let's see if your setup is conducive to detecting the subtle differences you're looking for. Build two small boxes with 4 RCA jacks in each. In one box, wire jack 1 to jack 2, then jack 3 to jack 4 (i.e., a two channel pass-through). In the second box, build a pair of passive RC step networks with a turnover frequency of approximately 1kHz and a step size of 0.3 dB (above the established threshold for detectability).
Have your friend switch boxes in and out. Can you identify straight versus EQ? If not, either your ears or your system do not have enough resolving power to detect well-established limits, so it's useless to try to detect things which are even tougher.
(as a side note, don't let your friend choose which one to put in during the test- have him use a random number generator or a coin flip ahead of time to determine the test order)
Let's see if your setup is conducive to detecting the subtle differences you're looking for. Build two small boxes with 4 RCA jacks in each. In one box, wire jack 1 to jack 2, then jack 3 to jack 4 (i.e., a two channel pass-through). In the second box, build a pair of passive RC step networks with a turnover frequency of approximately 1kHz and a step size of 0.3 dB (above the established threshold for detectability).
Have your friend switch boxes in and out. Can you identify straight versus EQ? If not, either your ears or your system do not have enough resolving power to detect well-established limits, so it's useless to try to detect things which are even tougher.
(as a side note, don't let your friend choose which one to put in during the test- have him use a random number generator or a coin flip ahead of time to determine the test order)
I don't think you've read the whole thing or probably there are some things that are not clear that caused you think that.
Apparently being able to identify differences and being able to correctly identify differences and allocate the specific sonic characters are different levels that need to be addressed. With a group of people doing ABX without ability to each take their time to listen the way they feel necessary will cause the whole group to show statistics similar to randum guessing. My friend actually recommended that if I did the tests the ABX way, the rates would have gone even higher. An additional issue is how loud you do the tests if the range is beyond the best performance level of the driver or the amp, then the capability to correctly identify is less. We had no way of lowering the volume during this test. Maybe next time when we have the time.
I encourage others that think they can hear a difference to to the test themselves and honestly see how the results are. I'm sure younger ears should be better.
Is anyone going for that $1M?
Apparently being able to identify differences and being able to correctly identify differences and allocate the specific sonic characters are different levels that need to be addressed. With a group of people doing ABX without ability to each take their time to listen the way they feel necessary will cause the whole group to show statistics similar to randum guessing. My friend actually recommended that if I did the tests the ABX way, the rates would have gone even higher. An additional issue is how loud you do the tests if the range is beyond the best performance level of the driver or the amp, then the capability to correctly identify is less. We had no way of lowering the volume during this test. Maybe next time when we have the time.
I encourage others that think they can hear a difference to to the test themselves and honestly see how the results are. I'm sure younger ears should be better.
Is anyone going for that $1M?
SY said:
I'm not sure how it differs as long as I don't know. How would you do that with three DUT instead of two?
SY said:
This is quite similar to the techniques used in the "Gold Ear Training Program" Have you used it?
First, you ran the test without a control- you need to first establish that the test/listener is capable of basic, well-established detection before trying the fancy stuff. That was the point I was trying to make with my suggested test.
Second, true randomization is necessary. Humans do NOT pick orders randomly, hard as they try (especially repeats- a random number generator might have the same thing come up three times in a row, whereas a human would never do that). And an advantage to having the trial order done in advance is that there's less cluing from non-verbal cues. Ideally, you'd not see or interact at all with the guy doing the switching, but second best is to make it as easy as possible for him to keep a poker face.
Second, true randomization is necessary. Humans do NOT pick orders randomly, hard as they try (especially repeats- a random number generator might have the same thing come up three times in a row, whereas a human would never do that). And an advantage to having the trial order done in advance is that there's less cluing from non-verbal cues. Ideally, you'd not see or interact at all with the guy doing the switching, but second best is to make it as easy as possible for him to keep a poker face.
This should cause some turbulence, but I believe if the cables don't measure the same, in human hearing ranges, you need to adjust for that before doing the test. For example if you can measure a capacitance on one, you should add capacitance on the other.
macgyver10 said:
I categorize them.
1. Silly things that are outright foolish. Things like the electrons falling out of the cable if it is bent too much, electrons having to go downhill vs up..
2. Silly things that are not so obvious, like piezoelectric effect on speaker wires, grain boundary.
3. Things that seem plausible at first glance, like motor generator, cable lifters, antistatic spray...but upon examination, have effects 5 to 10 orders of magnitude below audibility.
4. Things that seem plausible at first glance, then don't seem plausible at first calc blush, then, get ya thinkin..then, thinkin some more. These are the most fun for me, as it requires I learn something new to understand the problem.
5. Things that are completely taken out of context, like prop velocity, electron speed, reflection coeff. Real entities, and some even correlate to effects, but are not causation.
But yes, some cannot be categorized so easily.
tomtt said:
TTable, cd, comp at times..I have a simple 5.1 set, and do some PA stuff, nothing I'd recommend sitting in the sweetspot for to view a soundstage. Gross phase/image abnormalities I can "see", but I never have the time to tweak that. Instead, I just enjoy the artist at work.
Cheers, John
Sorry that I'm not trying to establish a procedure for just anyone, I designed it so that it is reasonalby difficult for me to identify the proper interconnects. Based on the statistics of three in a row is really unlikely in a sample of 10.SY said:
My friend could not even be seen, he was behind a 2.4MX2.4M wall. We arrange the material I selected in specific order so I just say "CD#, Track#, change cable" and he does it, and never responds. The reason for asking him to disconnect the cables anyway even if he does not change is to make sure the time taken is about the same and I hear things are being done. We don't see or interact with each other unless the test session is finished.
I really think people should just try for themselves. Just try to be as realistic as you can. It only took us about 4 hours from setup to finish.
- Status
- Not open for further replies.